Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Enhanced Transposition Cutoff

Author: Robert Hyatt

Date: 08:28:14 05/18/98

Go up one level in this thread


On May 18, 1998 at 09:49:40, Peter Fendrich wrote:

>Has anyone tried this?
>
>I found the following description at:
>http://www.xs4all.nl/~verhelst/chess/search.html#enhanced-transposition
>----------------------------------------------------------------
>There may be several successors causing a cutoff and we want to use the
>one with the smalles search tree. One idea that has been tried is to
>look at all successor positions and see if they are in the transposition
>table and cause a cutoff. If one such position is found, no further
>serach has to be done.
>This can save about 20-25% in total tree size.
>----------------------------------------------------------------
>
>
>//Peter

I tried it about 2 years ago...  It does two things:  (1) it will reduce
the size of the tree searched by about 10% according to the tests I ran
way back then.  (2) it breaks programs that generate moves in batches,
because you can't... you have to generate *all* the moves, and use each
move to produce a new hash signature that you use to probe the hash
table
to decide which move to try.  In Crafty, this cost me more than that 10%
saving in nodes searched.

If you always generate all the moves (Cray Blitz did because the moves
were vectorized) then this works, and we used it there with success.
But
if you defer move generation as I do in crafty, by generating only
captures,
and then trying killers before generating non-captures, you lose the
benefits of that.  A program like dark thought that "generates as
needed"
would probably slow down even more, assuming it still generates moves a
few at a time.




This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.